By: Dr. Loïc VERLINGUE, Physician-Researcher at the Léon Bérard Center & Dr. Tawhid CHTIOUI, Founding President of aivancity, the leading school for AI and dataa
Artificial intelligence represents one of the most promising technological advancements of our time, particularly in the field of healthcare. Through its potential, it could not only improve the quality of care but also redefine the relationship between patients and healthcare professionals. Its applications can help address contemporary challenges such as managing chronic diseases, reducing the administrative burden on healthcare providers, and improving access to care across the country.
But this revolution is not without its challenges: how can we ensure that these technologies remain at the service of people, that they strengthen the role of healthcare providers while empowering patients in their healthcare journey? How can we guarantee the ethical and responsible use of digital tools, and how can we transform medical practice without dehumanizing care?
At the heart of this transformation lies a fundamental question:
Can AI serve as a catalyst for humanizing the healthcare system while enabling significant scientific, economic, and organizational progress?
This reflection requires putting the patient back at the center, valuing healthcare providers, and rethinking training to support this transformation, all while addressing ethical and environmental concerns.
Empowering patients through AI
With AI, patients can take a leading role in their own health by actively participating in the management and monitoring of their condition. AI-based technologies offer powerful tools for democratizing access to information and personalizing care in unprecedented ways. For example, mobile apps now allow patients with chronic conditions, such as diabetes or hypertension, to track their daily progress using smart sensors. These devices provide them with accurate real-time data and tailored recommendations to help them manage their condition effectively. This type of technological support not only enhances their autonomy but also deepens their understanding of the mechanisms of their disease.
Some tools can even take into account a patient’s preferences, whether regarding their lifestyle, priorities, or concerns about certain treatments. It is now necessary to demonstrate long-term benefits in terms of patient and caregiver satisfaction, reduced complications, and lower healthcare costs. For example, a patient wishing to minimize hospitalizations could benefit from a tailored home treatment plan, thanks to telemedicine and remote monitoring tools. Chatbots—offering, for example, psychosocial support—and consultation summary tools have the potential to improve overall patient care, respecting individuals’ needs and expectations, while strengthening their trust in the healthcare system. AI has the potential to reconcile high technology with a human approach. To achieve this, there are many pitfalls to avoid. In particular, the risk of impairing communication with patients and among healthcare providers themselves. Medical and nursing common sense is likely the best guide, helping to maintain meaningful practices and fostering the vocations of future professionals.
Early and optimized diagnoses
In the fight against cancer, AI systems offer promising prospects. Capable of analyzing thousands of medical images very quickly, these technologies help detect abnormalities early on, at stages where prompt intervention can alter the course of the disease. This has been widely demonstrated in the automated analysis of mammograms under human supervision, for example. However, acquisition costs are sometimes high relative to the expected time savings. This capacity for in-depth analysis can enable accurate diagnoses, for example through the analysis of histopathological slides, in very specific situations. Multi-cancer early detection (MCED) tests go even further: they can predict the site of a cancer at a very early stage—sometimes even before it is visible on imaging—through a simple blood test. They are based on powerful AI algorithms and show encouraging preliminary results. Many questions remain regarding their clinical applications, particularly regarding the risk of false positives. It is in France’s best interest to participate more extensively in these evaluations.
Furthermore, a broad field is dedicated to replicating human performance in the personalization of care. Personalized medicine enables a better response to patients’ specific needs by taking into account their medical history, preferences, and personal circumstances, combined with data generated using cutting-edge techniques (such as high-throughput sequencing). This information is made accessible to patients, such as through personalized clinical trial search tools or the interpretation of molecular analyses of cancers. Ultimately, AI does more than just assist in care: it can empower patients and support them in a process of continuously improving their quality of life.
From reaction to prevention: a paradigm shift
Certain technological innovations can even promote preventive and sustainable health. For example, a patient with cardiovascular risk factors or at risk of adverse effects from chemotherapy can receive personalized alerts, along with specific recommendations for adjusting their lifestyle. By using real-time monitoring tools, these patients can track the impact of their choices on their health, whether through dietary changes, exercise programs, or tailored medication regimens.
Beyond individual benefits, this preventive approach also has significant economic implications. By reducing costs associated with treatments and hospitalizations, preventive AI can help ensure the sustainability of healthcare systems. For example, in the context of chronic diseases and cancers, predictive technologies are being developed to identify at-risk patients before complications arise, thereby reducing the need for emergency interventions. This proactive approach also improves patients’ quality of life by sparing them acute episodes or more invasive treatments. By shifting practices toward a preventive model, AI helps keep people healthy for as long as possible.
Treating each patient as a unique case
AI is used to develop new drugs at a very high throughput. The very first molecules generated by AI appear to have a very promising success rate in early-phase clinical trials. These molecules are therefore highly biologically targeted and optimized to minimize potential side effects. The number of new molecules to be evaluated in clinical trials continues to grow exponentially, and this trend is set to continue. Clinical trials provide cancer patients with access to cutting-edge and highly personalized care. They also have positive economic impacts, help anticipate future healthcare costs, and promote scientific influence in a highly competitive global environment. There is much to be gained by working on regulatory processes, aligning with European initiatives, and reducing the border effect, in order to continue attracting new molecules proposed by pharmaceutical companies using AI—including some French ones—and thereby maintain the quality of care provided to patients now and in the future.
Most of the applications presented, taken together, require the analysis of multiple types of health data (text, biological, molecular, imaging, etc.). Access to this individual data remains a challenge, requiring a regulatory framework that ensures data security while promoting the use of high-quality tools. We must also anticipate that the prescription of AI-based tools for healthcare support will also be personalized. For example, a conversational agent providing psychological support cannot be suitable for all patients indiscriminately. With the multitude of AI applications in healthcare, it is time to personalize their use in clinical practice, just as we do when prescribing a CT scan, a bronchoscopy, or a molecular analysis.
AI Driving New Therapeutic Strategies
AI enables the analysis of complex databases, transforming and combining existing information to enrich it and reveal new insights. It does not create knowledge per se, but it facilitates its discovery by highlighting correlations or trends that were previously difficult to identify. For example, an analysis of data from several hospitals revealed that patients treated with immunotherapy for cancer appear to have a lower incidence of secondary cancers (in other locations) after recovery. This observation, derived from existing data and interpreted using AI, led to the launch of an academic clinical trial, PREDOSTAR, exploring the use of “preventive” immunotherapy following curative treatment in cancer patients. This type of application illustrates how AI, within the context of academic research, can enrich and enhance knowledge, paving the way for new therapeutic options.
Training the Doctors of Tomorrow: Balancing Science and Empathy
To reap the full benefits of these advancements while avoiding risks, it is essential to train all generations of healthcare providers in these developments. This presents a wonderful opportunity to reflect on the importance of interpersonal relationships and communication in healthcare. Healthcare remains the setting where people discuss their personal health issues. It is crucial to foster these qualities in healthcare providers, such as active listening and emotional presence. Even though language models demonstrate surprising and useful capabilities for empathy, they will not be able to respond to all complex medical situations in the field.
The doctors of tomorrow will need to be trained to work alongside technological tools, enabling them to enhance their technical and multidisciplinary skills while emphasizing the importance of a human-centered approach in their practice. This evolution requires learning the basics of AI to understand the underlying data and algorithms, as well as applying this knowledge in real-world situations where patient interaction remains the top priority. This educational investment also helps maintain high-quality research in an ultra-competitive ecosystem.
Interdisciplinary collaboration must also become the norm. Healthcare professionals will need to work hand in hand with data scientists, engineers, and ethics experts to design solutions tailored to patients’ needs. This hybrid approach will maximize the benefits of AI while minimizing risks, such as algorithmic bias or data privacy issues. It also helps identify the challenges and human resources essential for deploying these systems in sensitive contexts, such as underserved populations or areas suffering from medical deserts.
Transforming Healthcare with AI: A Solid Foundation for the Future
The changes brought about by modern AI are part of a long history of improvements to the healthcare system driven by technological advancements. It can even enhance the way we think about and deliver care, both technically and on a human level. The keys to success are undoubtedly putting the patient at the center of our concerns and ensuring high-quality working conditions for a sufficient number of healthcare providers, all of which foster a virtuous cycle. This transformation cannot happen without significant investment in training, research, and interdisciplinary collaboration. On the horizon lies a system where healthcare providers, engineers, decision-makers, and patients jointly develop innovative solutions and evaluate their applications and impacts within their respective fields, taking into account societal, cultural, and environmental impacts. AI calls for paving the way toward better integration of health data at the European and global levels, to ensure population representativeness and to promote knowledge sharing and the acceleration of scientific discoveries.
This revolution, however, calls for greater ethical vigilance: ensuring equitable access to technologies, respecting data privacy, promoting AI systems under European control, and ensuring that the tools developed truly serve the interests of patients. By pushing beyond the current boundaries of innovation, we could envision a future where AI not only anticipates diseases but also contributes to enhancing the overall well-being of individuals. It is on this condition that AI will be able to fully realize its promise: to become a driver of humanization and excellence for the healthcare system of tomorrow.
